Post on 25-Apr-2023
NECK CIRCUMFERENCE AND CARDIOVASCULAR RISK | 161
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014 Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003 ORIGINAL | ORIGINAL
1 Universidade Federal de Viçosa, Centro de Ciências Biológicas e da Saúde, Departamento de Nutrição e Saúde. Av. PH Rolfs,s/n., Campus Universitário, 36570-000, Viçosa, MG, Brasil. Correspondência para/Correspondence to: VSS GONÇALVES.E-mail: <vivian.goncalves@ufv.br>.
2 Universidade Federal do Espírito Santo, Centro de Ciências Agrárias, Departamento de Farmácia e Nutrição. Alegre, ES, Brasil.
Sponsors: Conselho Nacional de Desenvolvimento Científico e Tecnológico (Process nº 485986/2011-6) and Fundação de Amparo àPesquisa do Estado de Minas Gerais (Process nº APQ-01618-10).
Neck circumference as predictor of excessbody fat and cardiovascular riskfactors in adolescents
Perímetro do pescoço como preditor de excesso
de gordura corporal e fatores de risco
cardiovascular em adolescentes
Vivian Siqueira Santos GONÇALVES1
Eliane Rodrigues de FARIA2
Sylvia do Carmo Castro FRANCESCHINI1
Silvia Eloiza PRIORE1
A B S T R A C T
Objective
To verify whether neck circumference can predict cardiovascular risk factors and excess body fat in adolescents.
Methods
This cross-sectional study included male and female adolescents aged 10 to 14 years from Viçosa, Minas Gerais.The following data were collected: anthropometric measurements, blood pressure, percentage of body fataccording to dual energy X-ray absorptiometry, and levels of fasting glucose, fasting insulin, triglycerides, totalcholesterol, high-density lipoprotein, and low-density lipoprotein. The anthropometric measurements wereused for calculating indices and assessing nutritional status. The receiver operating characteristic curve testedwhether neck circumference could predict cardiovascular risk. We also investigated how neck circumferencerelated to the study parameters. The significance level was set at 5% (p<0.05).
Results
A total of 260 adolescents were assessed of which 50.4% (n=131) were females, 20.4% (n=53) had excessbody weight according to the body mass index-for-age index, and 42.7% (n=111) had excess body fat. Bloodpressure (3.9%, n=10) and all biochemical parameters were affected, varying from 1.9% (n=5) for glucose to
162 | VSS GONÇALVES et al.
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
65% (n=169) for total cholesterol. Neck circumference correlated with body fat, waist circumference, hipcircumference, weight, height, body mass index, waist-to-height ratio, homeostasis model of assessment -insulin resistance, fasting insulin, blood pressure, and high-density lipoprotein (<0.05). It predicted excess bodyfat and changes in fasting insulin and blood pressure in both sexes, blood glucose and high-density lipoproteinin females, and triglycerides in males (area under the curve >0.5; p<0.05, varying from 0.610 for high-densitylipoprotein to 0.817 for blood pressure).
Conclusion
Neck circumference was capable of predicting excess fat and cardiovascular risk factors in adolescents.
Indexing terms: Adiposity. Adolescent. Cardiovascular diseases. Neck.Triage.
R E S U M O
Objetivo
Verificar a capacidade do perímetro do pescoço, predizer fatores de risco cardiovascular e excesso de gorduracorporal na triagem de adolescentes.
Métodos
Estudo transversal com adolescentes de 10 a 14 anos, de ambos os sexos, residentes em Viçosa, Minas Gerais.Foram realizadas avaliações de medidas antropométricas, pressão arterial, gordura corporal pelo AbsortometriaRadiológica de Dupla Energia, glicemia e insulinemia de jejum, triglicerídios, colesterol total e frações lipoproteínade alta densidade e lipoproteína de baixa densidade. As medidas antropométricas foram utilizadas para cálculode índices e avaliação do estado nutricional. O perímetro do pescoço foi testado como preditor de riscocardiovascular através da Curva ROC e verificada sua relação com os parâmetros avaliados. Adotou-se significânciade p<0,05.
Resultados
Foram avaliados 260 adolescentes: 50,4% (n=131) do sexo feminino, 20,4% (n=53) com excesso de peso peloíndice de massa corporal/Idade e 42,7% (n=111) com excesso de gordura corporal. Houve alterações em todosos parâmetros bioquímicos - com variação de 1,9% (n=5) para glicemia a 65% (n=169) para colesterol total -e na pressão arterial (3,9%, n=10). O perímetro do pescoço correlacionou-se com gordura corporal, perímetroda cintura, perímetro do quadril, peso, estatura, índice de massa corporal, relação cintura/estatura, homeostasismodel of assessment-insulin resistance, insulinemia, pressão arterial e lipoproteína de alta densidade (p<0,05),sendo capaz de predizer o excesso de gordura corporal e alterações na insulina de jejum e na pressão arterialpara ambos os sexos; glicemia e lipoproteína de alta densidade para o sexo feminino e triglicerídeos para omasculino (Area Under the Curve >0,5; p<0,05, com variação entre 0,610 para lipoproteína de alta densidadee 0,817 para pressão arterial).
Conclusão
O perímetro do pescoço foi capaz de predizer o excesso de gordura corporal e fatores de risco cardiovascularna população adolescente.
Termos de indexação: Adiposidade. Adolescente. Doenças cardiovasculares. Pescoço. Triagem.
I N T R O D U C T I O N
Adolescence is a period of physical,
behavioral, and emotional changes that marks thetransition between childhood and adulthood.During this phase, which chronologically lasts
from ages 10 to 19 years, individuals are
consolidating food habits, and so are morevulnerable both to nutritional deficiencies because
of the high amounts of energy and nutrientsneeded to meet the intense growth that occursduring this period, and to obesity because of the
excessive intake of energy-dense foods. Besidesfood, other factors such as smoking, alcohol
NECK CIRCUMFERENCE AND CARDIOVASCULAR RISK | 163
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014 Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
abuse, and inactivity contribute to adolescents’vulnerability to the risk factors for metabolicdisorders and cardiovascular diseases1,2.
According to the Programa de OrçamentoFamiliares (POF, Family Budget Survey) conductedby the Instituto Brasileiro de Geografia e Estatística(IBGE, Brazilian Institute of Geography andStatistics), between 2008 and 2009 roughly 20%of the adolescents living in metropolitan areas inBrazil were overweight3. Excess body weight,when due to the accumulation of body fat, isclosely related to the incidence of obesity andmany other diseases, so early detection isimportant4.
Neck circumference has been investigatedas a screening instrument for overweight individualsbecause it is easy to measure, inexpensive,noninvasive, and unlike waist circumference, itdoes not vary throughout the day. Moreover, neckcircumference correlates with many fat-relatedanthropometric measurements and cardiovascularrisk factors5-8.
Given that anthropometric assessment iseasy and practical and that early detection ofexcess body fat in adolescents is important, theobjective of this study was to verify the extent towhich neck circumference can predictcardiovascular risk factors and excess body fat inadolescent screenings.
M E T H O D S
This epidemiological, cross-sectional studyincluded male and female adolescents aged 10to 14 years and 11 months selected by simplerandom sampling in all public and private schoolsin the rural and urban areas of the municipalityof Viçosa, Minas Gerais, Brazil. The inclusioncriteria were: not taking medication that affectsthe metabolism of lipids and carbohydrates orblood pressure, and not having been diagnosedwith infections and/or acute inflammations andchronic Non-Communicable Diseases (NCD).
The sample size was determined byconsidering all the 10-to-14-year-old adolescents
living in the municipality in 2010, which totaled5,7529, and a rate of excess body fat of 17.5%found by another study of the local adolescentpopulation with similar ages10. The acceptablevariability was 5.0% with a 95% confidence level,totaling 214 adolescents. Considering a sampleloss of 20.0%, at least 257 volunteers would beneeded for the sample, 129 of each gender. Thetool StatCalc of the software Epi Info version 6.04was used for the cross-sectional studies in thecalculation.
Students were selected among those whosigned the Informed Consent Form, respectingthe proportion of students of each age in eachschool. When a participant refused to participatein any of the study stages or dropped out, anotherstudent was selected to replace him/her. Thirteenindividuals were excluded from the study becauseof the following reasons: taking anticonvulsants(1); inappropriate fasting (2); blood could not becollected for technical reasons (1); history ofmyelomeningocele (1); nose cancer (1); takinggrowth hormone (1); hormonal changes (3);taking dietary supplements and/or vitamins (3).
The project was approved by the HumanResearch Ethics Committee of the UniversidadeFederal de Viçosa (UFV) under Protocol number0140/2010. The adolescents and their guardianssigned an Informed Consent Form created asrecommended by Resolution 196/96 of theNational Health Council.
Anthropometric and body compositionassessment
Body weight was determined by anelectronic digital scale with a maximum capacityof 150 kg and accuracy of 50 g. Height wasdetermined by a portable stadiometer with amaximum length of 2.13 m and accuracy of0.1 cm. All measurements were taken twice, andthe average was used for the study. When thetwo height measurements differed by more than0.5 cm, a new measurement was taken. BodyMass Index (BMI)-for-age and height-for-age were
164 | VSS GONÇALVES et al.
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
calculated to characterize the population, usingthe World Health Organization’s (WHO) cut-offpoints (Z-scores) as reference11.
Waist circumference was measured by atwo-meter flexible, inelastic tape measure withan accuracy of 0.01 cm, making sure not tosqueeze the soft tissues, at the midpoint betweenthe last rib and the iliac crest12. The waist-to-heightratio was given by dividing the waistcircumference in cm by the height in centimetre.
The hip circumference was measured atthe gluteal region at the widest circumferencebetween the waist and the knees13. The waist-to-hip ratio was then calculated by dividing thewaist circumference in cm by the hip circumferencein centimetre.
Neck circumference was measured at itsmidpoint, except when the individual had apronounced Adam’s apple, in which case the neckcircumference was measured right below it. Theadolescents were advised to stare at the horizonand stand up straight14.
All anthropometric measurements weretaken by the same individual who had been dulytrained for the task.
The percentage of Body Fat (%BF) wasdetermined by the device Dual Energy X-rayAbsorptiometry (DEXA) (Lunar Prodigy AdvanceDXA System - analysis version: 13.31, GEHealthcare) at the Diagnostic Imaging sector ofthe Health Division of the UFV and analyzedaccording to adolescent-specific classificationproposed by Lohman15, who suggests 15% to25% of body fat for females and 10% to 20%for males.
Biochemical assessment
Participants’ blood (12 mL) was collectedby trained professionals after a 12-hour fast byvenipuncture using disposable syringes. Thelaboratory of clinical analyses of the HealthDivision of the UFV performed the biochemicalanalyses. The following were determined: totalcholesterol, triglycerides, High Density Lipoprotein-
cholesterol (HDL-c), Low Density Lipoprotein-cholesterol (LDL-c), fasting glucose, and fastinginsulin.
The blood was centrifuged by thecentrifuge Excelsa Model 206 BL for 10 minutesat 3,500 rpm right after collection but allowingenough time for the blood to coagulate. Totalcholesterol, HDL-c, and triglycerides wereautomatically determined by the device CobasMira Plus (Roche®) using an enzymatic method.LDL-c was given by the Friedewald equation.Nobody’s triglyceride level exceeded 400 mg/dL.Fasting glucose was automatically determined bythe device Cobas Mira Plus (Roche®) using the
glucose oxidase method. Fasting insulin wasdetermined by electrochemiluminescence.
Serum lipids were assessed as recommendedby the I Guidelines for Preventing Childhood and
Adolescent Atherosclerosis2. Fasting glucose levelswere classified as recommended by the SociedadeBrasileira de Diabetes16 (Brazilian Diabetes Society)
and fasting insulin as recommended by the So-ciedade Brasileira de Cardiologia2 (BrazilianCardiology Society). High and borderline values were
considered abnormal as follows: total cholesterol≥150 mg/dL, LDL-c ≥100 mg/dL, HDL-c <45 mg/dL,triglycerides ≥100 mg/dL, fasting glucose
≥100 mg/dL, and fasting insulin ≥15 µU/mL.
Insulin resistance was given by theHomeostasis Model Assessment - InsulinResistance Index (HOMA-IR) according to the
fasting insulin and glucose levels as follows:HOMA-IR=fasting insulin (µU/mL)xfasting glucose(mmol/L)/22.5 and interpreted as recommendedby the Sociedade Brasileira de Cardiologia 2.
Blood pressure
Blood pressure was measured by a digitalblood pressure monitor recommended by theSociedade Brasileira de Cardiologia and asrecommended by the manufacturer andinterpreted according to the VI BrazilianGuidelines on Hypertension. Both the systolic and
NECK CIRCUMFERENCE AND CARDIOVASCULAR RISK | 165
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014 Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
diastolic blood pressures were assessed forchanges17.
Blood pressure was classified according toheight percentiles for both genders as follows:normal when blood pressure <90th percentile(<120/80 mmHg); borderline when 90th percentile<blood pressure <95th percentile; high whenblood pressure >95th percentile. A blood pressureequal to or above 120/80 mm/Hg, even when<95th percentile, was considered borderline.Borderline and high blood pressures wereconsidered abnormal.
Statistical analyses
The data were entered twice in thespreadsheet Excel 2010 and analyzed by theprograms Epi Info 6.04, Sigma Stat for Windows3.5 and MedCalc 12.
The Kolmogorov-Smirnov test identifiedthe variables with normal distribution,determining which other tests should be used:Student’s t test or Mann-Whitney test, andPearson or Spearman correlations.
Receiver Operating Characteristic (ROC)curves were constructed to verify the ability ofneck circumference to predict biochemical, body
composition, and blood pressure changes in thestudy adolescents and to determine the respectivecut-off points for females and males. The AreasUnder the Curves (AUC) were calculated alongwith their respective 95% Confidence Intervals(95%CI). The null hypothesis was accepted whenAUC ≤0.50. The ROC curve quantitativelydescribes the performance of a diagnostic testwhose result can be treated as a continuousordinal or categorical variable depending on howit compares with the gold standard18.
The following were calculated: positiveand negative predictive values, sensitivity, andspecificity of neck circumference for predictingbiochemical, body composition, and bloodpressure changes associated with cardiovascularrisk.
The significance level was set at 5%(p<0.05).
R E S U L T S
Subject characteristics
A total of 260 adolescents participated inthe study. Their median age, weight, and BMIwere, respectively: 12.4 (10.1-14.9) years; 43.3
Table 1. Descriptive characteristics of the sample of 260 students aged 10 to 14 years from public and private schools from Viçosa (MG),
Brazil, 2011.
Gender
BMI-for-age
Underweight
Normal weight
Overweight
Obese
H-for-age
Stunted
Normal height
%BF
Low
Normal
Excessive
Characteristics
50.4
00.8
81.7
10.7
06.8
02.3
97.7
09.2
36.6
54.2
131
1
107
14
9
3
28
12
48
71
% n
Female
49.6
06.2
70.5
15.5
07.8
01.5
98.5
19.4
49.6
31.0
129
8
91
20
10
2
127
25
64
40
% n
Male
100.0
003.4
076.2
013.1
007.3
001.9
098.1
014.2
043.1
042.7
260
9
198
34
19
5
255
37
112
111
% n
Total
Note: H: Height; %BF: percentage of Body Fat; BMI: Body Mass Index.
166 | VSS GONÇALVES et al.
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
(25.1-92.8) kg; and 18.3 (12.5-33.8) kg/m2. Theirmean height were 153.6±10.1 cm. Theseparameters did not differ significantly betweenthe genders (p>0.05).
The median neck circumferences for thegeneral population, males, and females were,respectively: 30.2 (25.5-39.5) cm, 30.5 (25.7-39.5)cm, and 29.9 (25.5-34.5) cm. Males had greaterneck circumferences (p<0.05).
Table 1 shows other characteristics of thestudy sample.
Clinical and biochemical assessments
Table 2 shows the following results:biochemical tests, insulin resistance, and bloodpressure. Females had higher triglyceride levels,HOMA-IR, and diastolic blood pressure, whilemales had higher fasting insulin levels (p<0.05).The other study parameters did not varysignificantly between the genders (p>0.05).
Relationship between neck circumferenceand cardiovascular risk variables
Neck circumference presented correlationsthat varied from 0.51 to 0.88 with fat-related
parameters and fat distribution, such as waist and
hip circumferences, percentage of body fat, weight,
and BMI. It also correlated with waist-to-height
ratio, HOMA-IR, fasting insulin, and blood
pressure, with amplitudes varying from 0.29 to
0.62. The waist-to-hip ratio and HDL-c were also
positively correlated with neck circumference.
Regarding the lipid profile, neck circumference
correlated positively with triglycerides in females
and negatively with HDL-c in both genders. For
these parameters and stratum, neck circumference
may be used (Table 3).
Predictive ability of neck circumference
Table 4 shows the AUC, sensitivity,
specificity, positive predictive value, and negative
predictive value of neck circumference for the
study parameters with significant results. It also
shows the neck circumferences associated with
abnormal parameters (cut-off points).
The area under the curve varied from
0.616 for HDL-c to 0.807 for blood pressure.
Sensitivity varied from 61.9% for HDL-c to 80.0%
for blood pressure and fasting glucose. Specificity
Table 2. Measurements and rates of biochemical and blood pressure changes in students aged 10 to 14 years from public and private
schools in Viçosa (MG), Brazil, 2011.
Total cholesterol (mg/dL)
HDL-c (mg/dL)
LDL-c (mg/dL)
Triglycerides (mg/dL)
Fasting glucose (mg/dL)
Fasting insulin (mcU/mL)
HOMA-IR
SBP (mmHg)
DBP (mmHg)
Parameters
158.3±31.3
52 (24-100)
93.2 (42.2-165.4)
66 (28-206)**
86 (67-114)
8.4 (1.1-45.7)*
2.1 (0.4-7.2)*
96.5 (72.5-124.5)
60.9±6.3*
FemaleChanges
% n
71.8
23.7
41.9
18.3
02.3
17.6
17.6
02.3
94
31
55
24
3
23
23
3†
165.4±26.7
51 (32-117)
93 (2.9-149.2)
62 (7-194)**
86 (70-195)
9.9 (1.8-36.4)*
1.4 (0.2-10.6)*
96 (73.5-143)
58.7±7.8*
MaleChanges
% n
58.1
31.0
34.1
08.5
01.6
07.8
07.8
05.4
75
40
44
11
2
10
10
7†
161.9±29.2
52 (24-117)
93 (2.9-165.4)
64 (7-194)
86 (67-195)
6.3 (1.1-45.7)
1.8 (0.2-10.6)
96.5 (72.5-143)
59.8±7.1
TotalChanges
% n
65.0
27.3
38.1
13.5
01.9
12.7
12.7
03.9
169
71
99
35
5
33
33
10†
Note: **p<0.05; *p<0.001; Student’s t test or Mann-Whitney test; †prevalence of blood pressure changes considering both systolic and diastolic
blood pressures. Data with normal distribution were expressed as means ± standard deviations. Data with abnormal distribution were expressed
as medians and their respective amplitudes. The following were considered abnormal: total cholesterol ≥150 mg/dL, LDL-c ≥100mg/dL,
HDL-c <45 mg/dL, triglycerides ≥100 mg/dL, fasting glucose ≥100 mg/dL, and fasting insulin ≥15 mU/mL; blood pressure was considered abnormal
when above the 90th height percentile.
HDL-c: High Density Lipoprotein-cholesterol; LDL: Low Density Lipoprotein-cholesterol; HOMA-IR: Homeostasis Model Assessment - Insulin Resistance;
SBP: Systolic Blood Pressure; DBP: Diastolic Blood Pressure.
NECK CIRCUMFERENCE AND CARDIOVASCULAR RISK | 167
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014 Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
varied from 57.3% for fasting insulin to 78.4%for blood pressure.
In summary, the cut-off points of 28.8 cmfor girls and 30.4 cm for boys indicate thepossibility of parameter changes associated withcardiovascular risk.
D I S C U S S I O N
The study adolescents were in consonancewith the national reality3: 20.3% (n=53) hadexcess body weight, being overweight or obeseaccording to their BMI-for-age. Nevertheless, ahigher percentage was classified as having excessbody fat (42.7%, n=111), showing that manynormal-weight adolescents are already at risk forobesity-related metabolic disorders.
Changes in glucose metabolism, bloodpressure, and lipid profile were also found by otherstudies on adolescents of the same age group19-20.
Table 3. Correlation between neck circumference and
anthropometric and cardiovascular risk parameters of
students aged 10 to 14 years from public and private
schools in Viçosa (MG), Brazil, 2011.
Body fat
Waist circumference
Hip circumference
Weight
Body mass index
W-to-He ratio
W-to-H ratio
HOMA-IR
Fasting insulin
SBP
DBP
Total cholesterol
LDL-c
HDL-c
Triglycerides
-0.67**
-0.67**
-0.78**
-0.83**
-0.73**
-0.43**
-0.05**
-0.33***
-0.36**
-0.65**
-0.43**
-0.19**
-0.08**
-0.26**
-0.01**
-0.61**
-0.83**
-0.87**
-0.92**
-0.81**
-0.46**
-0.23**
-0.50**
-0.49**
-0.63**
-0.33**
-0.29**
-0.26**
-0.26**
-0.20**
-0.51**
-0.74**
-0.75**
-0.88**
-0.75**
-0.41**
-0.14**
-0.35**
-0.36**
-0.62**
-0.29**
-0.27**
-0.18**
-0.27**
-0.06**
Parameters Female (r) Male (r) Total (r)
Note: Pearson or Spearman correlation. *p<0.001; **p<0.05.
W-to-He ratio: Waist-to-Height ratio; W-to-H ratio: Waist-to-Hip ratio;HOMA-IR: Homeostasis Model Assessment - Insulin Resistance; SBP:Systolic Blood Pressure; DBP: Diastolic Blood Pressure; HDL-c: High Density
Lipoprotein-cholesterol; LDL-c: Low Density Lipoprotein-cholesterol.
Table 4. Ability of neck circumference to predict the cardiovascular risk of students aged 10 to 14 years from public and private schools
in Viçosa (MG), Brazil, 2011, according to the receiver operating characteristic curve.
Body fat
Total
Female
Male
Fasting glucose
Total
Female
Fasting insulin
Total
Female
Male
HDL-c
Total
Female
Triglycerides
Male
Blood pressure
Total
Female
Male
0.676***
0.711***
0.728***
0.682***
0.827***
0.703***
0.659***
0.902***
0.616***
0.645***
0.700***
0.807***
0.908***
0,747***
0.615-0.733
0.625-0.786
0.643-0.803
0.621-0.738
0.751-0.887
0.643-0.757
0.571-0.739
0.837-0.947
0.554-0.676
0.557-0.727
0.613-0.777
0.754-0.854
0.844-0.951
0.663-0.819
064.0
067.6
075.0
080.0
100.0
072.7
095.7
100.0
061.9
064.5
054.6
080.0
100.0
085.7
65.1
66.7
60.7
67.5
75.8
57.3
36.1
74.0
59.3
64.0
86.4
78.4
69.5
71.3
57.7
70.6
46.2
04.6
08.8
19.8
24.2
24.4
36.4
35.7
27.3
12.9
07.1
14.6
070.8
063.5
084.4
099.4
100.0
093.5
097.5
100.0
080.6
085.3
095.3
099.0
100.0
098.9
>30.30
>29.55
>30.40
>31.10
>31.00
>30.40
>28.80
>31.70
>30.40
>30.40
>33.50
>31.70
>30.80
>31.70
Parameters AUC 95%CI Sensit% Specif% PPV% NPP% COP
Note: *p<0.001; **p<0.05.
AUC: Area Under the Curve; SD: Standard Deviation; 95%CI: Confidence Interval of 95%; Sensit: Sensitivity; Specif: Specificity; PPV: PositivePredictive Value; NPP: Negative Predictive Value; COP: Cut-Off Point; HDL: High Density Lipoprotein-cholesterol.
168 | VSS GONÇALVES et al.
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
Chaves et al.19 corroborate the presentfindings in their study of a similar population:21.0% were overweight or obese and 54.2%,26.7%, 25.8%, and 20.0% presented changesin their total cholesterol, LDL-c, HDL-c, andtriglyceride levels, respectively. They also foundprevalences of hyperinsulinemia and hyperglycemiaof 8.3% and 1.7%, respectively. In both studies,the worst results were for total cholesterol andthe best, for fasting glucose.
The results reinforce the need of identifyingthis population early to develop an effectiveintervention. Such identification requires practicaland rapid screening methods with good accuracyand sensitivity. Waist circumference has been usedfor this purpose and relates to cardiovascular riskfactors both in Brazilian and other adolescents21-24.
However, waist circumference variesthroughout the day and its measurement mayabash adolescents because of the body changesthey experience during puberty. In this sense, neckcircumference is an option and its use is beinginvestigated in different ethnicities and agegroups8, 25-27.
In the present study, neck circumferencewas capable of predicting excess body fat for bothgenders (AUC >0.7) and correlated with themeasurements and indices used for identifyingfat deposits, such as hip circumference, waist-to-height ratio, and waist circumference. Neckcircumference also correlated with weight andwas associated with BMI, proving to be a usefulinstrument.
Hatipoglu et al.6 assessed healthy andobese Turkish children and adolescents aged 6 to18 years and also found neck circumference tocorrelate with BMI and waist circumference,considering it a good supplementary instrumentfor detecting overweight and obesity. Nafiu et al.7
found that these variables, in addition to bodyweight, also correlated with neck circumferencein a population of the same age bracket in theUnited States of America, stating that neckcircumference can be a good predictor of excessweight.
Although neck circumference did notcorrelate with blood glucose level, it predictedblood glucose changes in females quite well.Meanwhile, hyperinsulinemia, even by itself, hasbeen pointed out as a good indicator of insulinresistance in adolescents24, and neck circumferencehas shown good sensitivity for predictinghyperinsulinemia in female and male adolescents.Given that neck circumference has also correlatedwith HOMA-IR, there is evidence that it may beuseful for screening adolescents with high bloodglucose.
Studies that investigated the relationshipbetween neck circumference and parametersassociated with impaired glucose metabolism inadolescents were not found, but Vallianou et al.25,Laakso et al.28, Ben-Noun & Laor29 and Yang etal.30 found neck circumference to be associatedwith insulin resistance indicators in adults.
Neck circumference also indicated bloodpressure changes well, correlating with bothsystolic and diastolic blood pressure. High AUC(0.807) and sensitivity (80%) show that neckcircumference could identify a large number ofadolescents at risk.
Ben-Noun & Laor5,29 found neckcircumference to correlate with blood pressure inadult Israelis and an association between thechanges that occurred in both parameters overtime. Studies with adolescents were not found.
Even by itself, high blood pressure is alinear and continuous risk factor for cardiovasculardiseases. In population-based studies, bloodpressure is related with the risk of death anddisease17. Faria et al.20 found a 5.0% prevalenceof high blood pressure in adolescents, and Pintoet al.31 found that 5.6% and 9.9% of theadolescents aged 10 to 14 years from Bahia,Brazil, were hypertensive and prehypertensive,respectively.
Lauer & Clarck32 found that blood pressureand excess adiposity during adolescencecorrelated with blood pressure during adulthoodin a study that followed 2,445 individuals. Thisfact restates the importance of neck circumference:
NECK CIRCUMFERENCE AND CARDIOVASCULAR RISK | 169
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014 Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
not only it is associated with blood pressure andcapable of predicting blood pressure changes, butit is also related to body fat.
While neck circumference was notassociated with total cholesterol and LDL-c, itcorrelated with HDL-c and predicted low HDL-cin females. On the other hand, neck circumferencecorrelated with triglycerides and predictedtriglyceride changes only in males.
Gonçalves et al.33 studied a populationwith similar ages and found significantprevalences of low HDL-c in boys (26.9%) andgirls (30.6%) and high triglycerides in boys(18.6%) and girls (25.6%). Gontijo et al.34
corroborated these findings in another study withadolescents aged 10 to 19 years. The high numberof dyslipidemic adolescents reinforces theimportance of identifying them for earlyintervention.
Ben-Noun & Laor5 found neck circumferenceto be related to serum lipid levels in adults.Vallianou et al.25 found neck circumference tocorrelate with HDL-c and triglyceride levels inadults, corroborating the present study.
The Sociedade Brasileira de Cardiologia2
recommends that the identification and treatmentof dyslipidemic children and adolescents shouldbe done early to prevent atherosclerosis, amongothers. Simple and easy-to-use methods, such asmeasuring neck circumference, may help toidentify these individuals.
The study adolescents already presentchanges in cardiovascular risk parameters,suggesting the importance of identifying andmonitoring this population from the beginningof this life phase.
Protocols should be created for primarycare and other units of the health care networksthat use simple, accurately reproducible
measurements, such as neck circumference, aconvenient and accessible alternative. However,the diagnosis must be confirmed by more specificmethods, ones that can better measure the studyparameters.
The suggestion of specific cut-off pointsfor the neck circumference of adolescents can beconsidered a strength of the present study, sincecut-off points enable the use of this measurement.However, a study limitation is the absence of datafor the last years of adolescence, preventing thedetermination of references for this entire lifephase.
Finally, the prediction of changes wasgenerally more sensitive than specific. Since thesensitivity of screening instruments is moreimportant than specificity for the nutritionaldiagnosis, we reiterate that neck circumferenceis a good alternative for this end.
C O N C L U S I O N
Neck circumference was capable of
predicting excess body fat and cardiovascular riskfactors in adolescents aged 10 to 14 years. It iseasy and quick to measure, noninvasive, and
inexpensive, and can be used as a screening toolin population-based studies.
Many changes have been found in thestudy parameters, demonstrating the importanceof health care programs for adolescents thatpromote their health and prevent future diseases.
A C K N O W L E D G M E N T S
The authors thank the MSc. Daniela Alves Silvaand MSc. Franciane Rocha de Faria for their valuablecontribution to data collection and Dr. Sônia MachadoRocha Ribeiro for reading and making suggestions to
the last version of the manuscript.
C O N T R I B U T O R S
All authors participated in all phases of the
article’s research.
R E F E R E N C E S
1. World Health Organization. Nutrition in adolescence:Issues and challenges for the health sector: Issues
170 | VSS GONÇALVES et al.
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
in adolescent health and development. Geneva:WHO; 2005.
2. Sociedade Brasileira de Cardiologia. I Diretriz deprevenção da aterosclerose na infância e na adoles-cência. Arq Bras Cardiol. 2005; 85(Supl 6):1-36.doi: 10.1590/S0066-782X2005002500001
3. Instituto Brasileiro de Geografia e Estatística. Pes-quisa de orçamentos familiares 2008-2009: despe-sas, rendimentos e condições de vida. Rio de Ja-neiro: IBGE; 2010.
4. Rech CR, Cordeiro BA, Petrosky EL, Vascolcelos FAG.Utlização da espessura de dobras cutâneas para aestimativa da gordura corporal em idosos. Rev Nutr.2010; 23(1):17-26. doi: 10.1590/S1415-52732010000100003
5. Ben-Noun L, Laor A. Relationship between changesin neck circumference and changes in bloodpressure. Am J Hypertens. 2004; 17(5 Pt 1):409-14.doi: 10.1016/j.amjhyper.2004.02.005
6. Hatipoglu N, Mazicioglu MM, Kurtoglu S, KendirciM. Neck circumference: An additional tool ofscreening overweight and obesity in childhood. EurJ Pediatr. 2010; 169(6):733-9. doi: 10.1007/s00431-009-1104-z
7. Nafiu OO, Burke C, Lee J, Voepel-Lewis T, MalviyaS, Tremper KK. Neck circumference as a screeningmeasure for identifying children with high bodymass index. Pediatrics. 2010; 126(2):306-10. doi:10.1542/peds.2010-0242
8. Mazicioglu MM, Kurtoglu S, Ozturk A, HatipogluN, Cicik B, Ustunbas HB. Percentiles and meanvalues for neck circumference in Turkish childrenaged 6-18 years. Acta Paediatr. 2010; 99(12):1847-53. doi: 10.1111/j.1651-2227.2010.01949.x
9. Brasil. Ministério da Saúde. Datasus. Brasília: Minis-tério da Saúde; 2010 [acesso 2010 nov 10]. Dispo-nível em: <http://tabnet.datasus.gov.br/cgi/deftohtm.exe?ibge/cnv/popmg.def>.
10. Chaves OC, Franceschini SCC, Ribeiro SMR,Sant’Ana LFR, Faria CG, Priore SE. Comparison ofthe biochemical, anthropometric and bodycomposition variables between adolescents from10 to 13 years old and their parents. Nutr Hosp.2012; 27(4):1127-33. doi: 10.3305/nh.2012.27.4.5832
11. World Health Organization, de Onis M, OnyangoAW, Borghi E, Siyam A, Nishida C, et al.Development of a WHO growth reference forschool-aged children and adolescents. Bull WHO.2007; 85(9):660-7.
12. Cameron N. The measurement of human growth.Austrália: Croom-Helm; 1984.
13. Heyward VH, Stolarczyk LM. Método antropo-métrico. In: Heyward VH, Stolarczyk LM. Avaliação
da composição corporal aplicada. São Paulo:Manole; 2000.
14. Ben-Noun L, Sohar E, Laor A. Neck circumferenceas a simple screening measure for identifyingoverweight and obese patients. Obes Res. 2001;9(1):470-8.
15. Lohman TG. Assessing fat distribuition. Advancesin body composition assessment: Current issues inexercise science. Illinois: Human Kinects; 1992.
16. Sociedade Brasileira de Diabetes. Diretrizes da So-ciedade Brasileira de Diabetes. 3ª ed. Itapevi: AraújoSilva Farmacêutica; 2009.
17. Sociedade Brasileira de Cardiologia, SociedadeBrasileira de Hipertensão, Sociedade Brasileira deNefrologia. VI Diretrizes brasileiras de hipertensão.Arq Bras Cardiol. 2010; 95(Supl 1):1-51.
18. Martinez EZ, Louzada-Neto F, Pereira BB. A curvaROC para testes diagnósticos. Cad Saúde Colet.2003; 11(1):7-31.
19. Chaves OC, Franceschini SCC, Ribeiro SMR,Sant’Ana LFR, Faria CG, Priore SE. Anthropometricand biochemical parameters in adolescents andtheir relationship with eating habits and householdfood availability. Nutr Hosp. 2013; 28(4):1352-6.
20. Faria ER, Franceschini SCC, Peluzio MCG, Sant’AnaLFR, Priore SE. Correlação entre variáveis de com-posição corporal e metabólica em adolescentes dosexo feminino. Arq Bras Cardiol. 2009; 93(2):119-27.
21. Ranjani H, Sonya J, Anjana RM, Mohan V.Prevalence of Glucose Intolerance Among Childrenand Adolescents in Urban South India (ORANGE-2). Diabetes Technol Ther. 2012; 15(1):13-9. doi: 10.1089/dia.2012.0236
22. Spolidoro JV, Pitrez FM, Vargas LT, Santana JC, PitrezE, Hauschild JA, et al. Waist circumference inchildren and adolescents correlate with metabolicsyndrome and fat deposits in young adults. ClinNutr. 2013; 32(1):92-7. doi: 10.1016/j.clnu.2012.05.020
23. Lu X, Shi P, Luo CY, Zhou YF, Yu HT, Guo CY, et al.Prevalence of hypertension in overweight and obesechildren from a large school-based population inShanghai, China. BMC Public Health. 2013;13(1):13-24. doi: 10.1186/1471-2458-13-24
24. Acosta García E, Carías D, Páez Valery M, NaddafG, Domínguez Z. Exceso de peso, resistencia a lainsulina y dislipidemia en adolescentes. ActaBioquím Clín Latinoam. 2012; 46(3):365-73.
25. Vallianou NG, Evangelopoulos AA, Bountziouka V,Vogiatzakis ED, Bonou MS, Barbetseas J, et al. Neckcircumference is correlated with triglycerides andinversely related with HDL cholesterol beyond BMIand waist circumference. Diabetes Metab Res Rev.2013; 29(1):90-7. doi: 10.1002/dmrr.2369
NECK CIRCUMFERENCE AND CARDIOVASCULAR RISK | 171
Rev. Nutr., Campinas, 27(2):161-171, mar./abr., 2014 Revista de Nutrição
http://dx.doi.org/10.1590/1415-52732014000200003
26. Tibana RA, Teixeira TG, Farias DLD, Silva ADO,Madrid B, Vieira A, et al. Relação da circunferênciado pescoço com a força muscular relativa e osfatores de risco cardiovascular em mulheressedentárias. Einstein. 2012; 10(3):329-34. doi:10.1590/S1679-45082012000300013
27. Fitch KV, Stanley TL, Looby SE, Rope AM, GrinspoonSK. Relationship between neck circumference andcardiometabolic parameters in HIV-infected andnon-HIV-infected adults. Diabetes Care. 2012;34(4):1026-31. doi: 10.2337/dc10-1983
28. Laakso M, Matilainen V, Kiukaanniemi KS.Association of neck circumference with insulinresistance-related factors. Int J Obes. 2002;26(6):873-75. doi: 10.1038=sj.ijo.0802002
29. Ben-Noun L, Laor A. Relationship of neckcircumference to cardiovascular risk factors. ObesRes. 2003; 11(2):226-31.
30. Yang L, Samarasinghe YP, Kane P, Amiel SA, AylwinSJB. Visceral adiposity is closely correlated with neckcircumference and represents a significant indicatorof insulin resistance in WHO grade III obesity. ClinEndocrinol. 2010; 73(2):197-200. doi: 10.1111/j.1365-2265.2009.03772.x
31. Pinto SL, Silva RCR, Priore SE, Assis AMO, Pinto EJ.Prevalência de pré-hipertensão e de hipertensãoarterial e avaliação de fatores associados emcrianças e adolescentes de escolas públicas deSalvador, Bahia, Brasil. Cad Saúde Pública. 2011;27(6):1065-76. doi: 10.1590/S0102-311X2011000600004
32. Lauer RM, Clarke WR. Childhood risk factors forhigh adult blood pressure: The muscatine study.Pediatrics.1989; 84(4):633-41.
33. Gonçalves VSS, Chaves, OC, Ribeiro SMR, Sant´AnaLFR, Franceschini SCC, Priore, SE. Disponibilidadedomiciliar de lipídeos para consumo e sua relaçãocom os lipídeos séricos de adolescentes. Rev PaulPediatr. 2012; 30(2):229-36. doi: 10.1590/S0103-05822012000200012
34. Gontijo CA, Faria ER, Oliveira RMS, Priore SE.Síndrome metabólica em adolescentes atendidosem programa de saúde de Viçosa-MG. Rev BrasCardiol. 2010; 23(6):324-33.
Received on: 6/11/2013Final version on: 1/7/2014Approved on: 1/22/2014